Grand Rounds Archive

Grand Rounds December 13, 2024: Home Blood Pressure Telemonitoring and Nurse Case Management in Black and Hispanic Patients With Stroke: A Randomized Clinical Trial (Gbenga Ogedegbe, MD, MPH, FACP)

Posted on by Sophia Ramirez

Speaker

Gbenga Ogedegbe, MD, MPH, FACP
Dr. Adolph & Margaret Berger Professor of Population Health
NYU Grossman School of Medicine
Director, Institute for Excellence in Health Equity (IEHE)
NYU Langone Health

Keywords

Hypertension; Racial Disparities; Case Management; Telemonitoring

Key Points

  • There are significant racial disparities when it comes to stroke outcomes in the U.S., with Black and Hispanic populations experiencing poorer outcomes than Caucasian populations. A major predictor for these disparities is hypertension (HTN); controlling HTN is key to secondary stroke prevention.
  • Home Blood Pressure Telemonitoring (HBPTM) and Nurse Case Management (NCM) have proven efficacy in addressing multilevel barriers to HTN control. However, their effectiveness remains untested in stroke patients, and their implementation is sub-optimal in Black and Hispanic patients.
  • The research team looked at the comparative effectiveness of HBPTM alone versus HBPTM plus telephone-based NCM when administered to Black and Hispanic stroke survivors with uncontrolled hypertension. They measured within-patient change in systolic blood pressure (SBP) from baseline to 12 months and the rate of recurrent stroke at 24 months.
  • This was a multisite, practice-based, comparative effectiveness randomized clinical trial conducted at six public hospitals and three academic medical centers in New York City. About 225 patients were randomized to each of the two arms.
  • The study population was defined by low-income, minoritized communities with significant comorbidity. This is a departure from much of the existing literature, Dr. Ogedegbe noted, in which patient populations tend to be more insured, more highly educated, and have higher employment rates.
  • The research team found that both groups experienced significant SBP reduction at both the six-month and 12-month mark. The telehealth intervention that combined HBPTM with NCM led to greater SBP reduction than HBPTM alone. There was no significant difference in the rate of recurrent stroke between the two groups across a 24-month period.
  • These findings provide strong empirical evidence for widespread implementation in low-income stroke survivors with multiple comorbidities. Policymakers can use this evidence to implement these strategies in minority patients with stroke and uncontrolled hypertension.
  • Implementation challenges will include coverage of HBPM, which varies among private payers and Medicaid programs; clinical support services for use; access to the internet, to transmit data for virtual care; and integration of data into electronic medical records.

Discussion Themes

To effectively address inequities in hypertension, researchers must partner with policymakers, payers, and community-based organizations ahead of time, with a focus on team-based care.

Sustainment can be an Achilles Heel for embedded trials, as programs that have been shown to be successful aren’t always covered by insurance. Chronic disease management with team-based care is a good step that practices can invest in in the interim.

The fidelity to the NCM intervention, measured by the participation in 20 NCM calls and the compliance with HBPTM, was around 68% – a decent figure given the challenges of conducting a trial in a real-world setting, Dr. Ogedegbe noted.

Front-line doctors were instrumental in helping the research teams onboard the sites and think through recruitment.

Grand Rounds December 6, 2024: Opportunities and Challenges in the Use of Large Language Models for Post-Marketing Surveillance of Medical Products (Michael E. Matheny, MD, MS, MPH)

Posted on by Sophia Ramirez

Speaker

Michael E. Matheny, MD, MS, MPH
Director, Center for Improving the Public’s Health Through Informatics
Professor of Biomedical Informatics, Biostatistics, and Medicine
Vanderbilt University Medical Center
Staff Scientist, Geriatrics Research Education and Clinical Care Service
Associate Director, VA ORD VINCI
Tennessee Valley Healthcare System VA

Keywords

Artificial Intelligence; Large Language Models; Surveillance; Medical Products

Key Points

  • Increasingly, leaders in many disciplines are finding new applications for Artificial Intelligence (AI). Within healthcare, this technology is being used to support clinical decision-making; imaging processing; drug discovery; clinical trials; and as Ambient and Autonomous AI.
  • Large Language Models (LLMs) are a subset of generative AI. Since 2012, LLMs have emerged as a promising new technology with rapid growth, evolution of capacity and reach, and many potential applications in healthcare and clinical research.
  • There is significant interest in using LLMs to assist with patient trial matching, clinical trial planning, and the development of trial protocols and consent documents.
  • Another key area that LLMs could provide support in is medical product safety surveillance, with potential applications in adverse event detection, probabilistic phenotyping, and information synthesis.
  • The post-marketing surveillance space utilizes an ecosystem of healthcare data, imaging, radiology reports, insurance, structured data, medical literature, and social media. These sources could be integrated to conduct LLM reasoning and extractions.
  • Key challenges in safe and effective use of LLMs for this purpose include the lack of evaluation for medical product surveillance, the complexities of prompt engineering, hallucination risk (i.e., false positives), and the fact that evolving models over time challenge stable performance estimates.

Discussion Themes

 A segment of the clinical workforce could be trained to be “super users,” partnering with development teams in order to make sure that these tools are working appropriately in a clinical environment.

There is substantial interest in using LLMs to support clinical decision-making. However, studies have shown that the quality of the AI output can influence the performance of the clinicians. Especially in high-risk clinical environments, any drift in those algorithms could result in adverse clinical outcomes. The life cycle approach to conceptualization, development, implementation, surveillance, and maintenance will be necessary to achieve and maintain performance.

Grand Rounds November 22, 2024: Tranexamic Acid Versus Placebo to Reduce Perioperative Blood Transfusion in Patients Undergoing Liver Resection: HeLiX Trial (Paul J. Karanicolas, MD, PhD, FRCSC)

Posted on by Sophia Ramirez

Speaker

Paul J. Karanicolas, MD, PhD, FRCSC
Associate Professor of Emergency Medicine and Epidemiology
The Edmond Odette Cancer Centre @ Sunnybrook Health Sciences Centre
Professor of Surgery
University of Toronto

Keywords

Liver Cancer; Liver Resection; Blood Transfusions

Key Points

  • Liver resection is a major operation. Though it’s become safer over the past several decades, it requires blood transfusion in 20-40% of patients. Bleeding and blood transfusion are associated with substantial post-operative morbidity and an increase in mortality, re-operation, and re-admission rates.
  • A 2012 meta-analysis found that tranexamic acid (TXA) use in surgery was associated with lower blood transfusion rates. However, most of the trials included in the analysis involved cardiac and orthopedic surgery. There was limited evidence to suggest that TXA offered similar protective effects for liver resection patients.
  • The research team designed a pragmatic trial of adult patients undergoing cancer-related liver resection. Their primary outcome was the receipt of blood transfusion over the first seven days during or following surgery.
  • They randomized 1,384 participants to either a TXA arm or a placebo arm over eight years. They found that administration of TXA had no impact on blood transfusion, interoperative bleeding, or total blood loss; however, it increased the incidence of any surgical complications and major adverse events, with a possible increase in venous thromboembolism (VTE).
  • These findings were discordant with most existing evidence. Possible explanations included chance (unlikely, given the narrow confidence interval); a differing mechanism of bleeding in liver resection patients; and a difference in the dose or timing of TXA (also unlikely).
  • Based on this trial, the study team concluded that 1) TXA does not reduce blood transfusion or bleeding in patients undergoing liver resection; 2) TXA increases adverse events in this context with a signal towards increased VTE; and 3) TXA should not be used routinely in patients undergoing liver resection.
  • More work is necessary to determine whether TXA is beneficial and safe in other types of major oncologic surgery.

Discussion Themes

After some debate, the research team designed their trial to include all liver resection patients as opposed to only patients receiving major liver resections. Only 1/3 of patients have a major operation, and all liver resection patients have at least a 15% risk for bleeding and blood transfusion.

They landed on receipt of blood transfusion as the primary outcome because of the strong association between transfusion and other patient-important outcomes and due to the limited nature of blood transfusions as a resource.

Dr. Karanicolas reflected that the team should have collected adverse events data differently, grouping the complications into more meaningful categories; currently, they are limited in their ability to interpret the complications data.

The study team intentionally did not build an efficacy stopping rule into the trial due to concerns about bias. Having seen the trial through to the end, they had stronger evidence when they arrived at their surprising conclusion.

Grand Rounds November 15, 2024: Pragmatic Randomized Trial of Smartphone-Based Nudges to Reduce Distracted Driving Among US Auto Insurance Customers (M. Kit Delgado, MD, MS)

Posted on by Sophia Ramirez

Speaker

M. Kit Delgado, MD, MS
Associate Professor of Emergency Medicine and Epidemiology
Faculty Director, Penn Medicine Nudge Unit

Keywords

Distracted Driving; Auto Insurance; Smartphones

Key Points

  • Distracted driving leads to over 800,000 crashes, over 400,000 injuries, and over 3,000 deaths per year. Though most people understand the dangers of texting while driving, it doesn’t always stop them from engaging with their cell phones while on the road.
  • Dr. Delgado applied a heuristic model of human decision-making to explain this disparity. Using a case example from his time working in an Emergency Department, he illustrated how status quo bias, automaticity, present bias, recency bias, overconfidence bias, and social distance/norms all contributed to the failure of self-control and the resulting crash.
  • The research team partnered with Progressive Auto Insurance to determine whether a usage-based insurance program, redesigned with behavioral economic insights, reduces handheld phone use while driving. They tracked active cell phone usage using telematics.
  • The study had six arms: business as usual; weekly social comparison feedback (WSCF); a delayed standard financial incentive (SFI); WSCF + delayed SFI; WSCF + reframed SFI; and WSCF + doubled, reframed SFI. The delayed SFI was dispensed at the end of the intervention whereas the reframed SFI was dispensed weekly; the total amount of cash was the same.
  • Drivers in the fourth arm (WSCF + delayed SFI) saw a 15% reduction in active cell phone usage; drivers in the fifth arm (WSCF + reframed SFI) saw a 21% reduction. Doubling the financial incentive did not further improve the outcome. Notably, neither WCSF nor the SFI alone led to a significant decrease in active cell phone use when compared to the control.
  • For most participants, the effects of the intervention waned after the intervention period ended. However, there were a few positive outliers that maintained lower rates of active cell phone usage in the post-intervention period. The study team interviewed the outliers and identified a few key situational strategies that could help maintain long-term effects.
  • The study team designed a second trial with interventions that incrementally incorporated takeaways from the interviews, including the provision of a phone mount, access to commitment + habit-building tips, and gamification. This stacking design was intended to make it easier to do the right thing and provide motivation for incremental improvements.
  • They found that adding behaviorally-designed gamification, the participant’s level increasing or decreasing based on goal attainment and leaderboard competition, led to a 20% reduction in active phone use. The addition of a modest financial incentive on top of the gamification arm led to a 28% reduction. This time, the effects were sustained after the interventions ended.

Discussion Themes

The advantages of working with an industry sponsor include reach, ability to scale, and implementation opportunity. The downsides include limitations on what you can test, lengthy contracting processes, and some constraints around what data can be published. Dr. Delgado noted that, for this kind of work, the pros outweighed the cons.

The post-intervention evaluation period ranged from four to eight weeks due to limitations around enrollment and a maximum observation period.

One challenge was distinguishing between passenger trips and driver trips – a driver may let a passenger use their phone while driving, for example. They gave participants the option to retroactively reclassify trips and found that the reclassification rate was about 1%, even after incentives entered the picture.

Grand Rounds November 8, 2024: The Effect of Cash Benefits on Health Care Utilization and Health: A Randomized Study of an Income Support (Sumit Agarwal, MD, MPH, PhD)

Posted on by Sophia Ramirez

Speaker

Sumit Agarwal, MD, MPH, PhD
Assistant Professor
Division of General Medicine
University of Michigan

Keywords

Income Support; Cash Benefits; Low-Income Population

Key Points

  • Low-income patients face several barriers to care and achieving better health. Co-pays prevent people from filling prescriptions; transportation barriers prevent patients from attending appointments; emergency departments take the place of primary care; and people cope with anxiety, depression, and/or substance use amid economic vulnerability.
  • The research team sought to understand the effect of income support on health care utilization and health. They conducted a randomized controlled study of a cash benefit program led and implemented by the government of Chelsea, Massachusetts.
  • Several features of the intervention made it uniquely suited to understanding the effect of income: it was recurring; it was a large amount relative to the baseline income in the community, which was around $1,500 per month; it was unconditional; there were no restrictions on how the money could be spent; and it was randomized.
  • The research team assembled electronic health record (EHR) data from three major health systems in the Boston area. Participants were linked to their EHR records. Together, this data likely accounts for 77% of all acute care and 78% of all outpatient visits among Chelsea residents.
  • Their primary outcome was the number of emergency department (ED) visits. Their secondary outcomes included ED visits by type; outpatient use; clinical measures; and COVID vaccination.
  • The research team found a 27% reduction in emergency department visits, including hospitalizations, and a 21% increase in subspeciality care. There was no change in clinical measures nor in COVID vaccination rates.
  • These findings raised two questions: 1) Were people utilizing the emergency room less because they were increasingly able to access and utilize outpatient care? 2) Given the decrease in emergency department use and hospitalizations, are there potential cost savings here?
  • Dr. Agarwal suggested that substitution was likely not driving the decrease in emergency department utilization, noting that there was no change in the rate of primary care or urgent care visits, nor in prescriptions; the decrease in emergency department visits and increase in subspecialty care appeared in two distinct groups of people.
  • The study team estimated that the decrease in acute care utilization resulted in savings of $450/person for the healthcare system. When discussing the cost of running a cash benefit program, Dr. Agarwal noted, it is important to acknowledge potential offsets as well.

Discussion Themes

The collaboration with the city of Chelsea simplified funding; the road to IRB approval; and obtaining community engagement and buy-in.

Cash transfer programs in the U.S. tend to fall in one of two categories: Programs that target certain low-income populations, i.e. families with children, and, increasingly, programs that target patients with substantial barriers to healthcare, i.e. pediatric cancer patients. Both will yield insights into how we can target a cash transfer intervention more broadly.

The effects were sustained in the months after the nine-month intervention period, albeit at slightly lower levels. However, it’s unlikely that this effect will be maintained in the long term; a one-time, brief cash infusion probably cannot overcome the cumulative disadvantages driving health disparities.

About 70% of the funds were spent at places where food was the primary product, namely grocery stores and markets. Roughly 20% were spent on retail; 4% on utilities; 1% on transportation; and 0.4% on alcohol and smoking. This represents a counterpoint to the concern often expressed about cash transfer programs: that the funds will be spent on tobacco and alcohol.

Grand Rounds November 1, 2024: Congenital Heart Initiative: Redefining Outcomes and Navigation to Adult Centered Care (CHI-RON) Study (Thomas W. Carton, PhD, MS; Anitha S. John, MD, PhD)

Posted on by Sophia Ramirez

Speakers

Anitha S. John, MD, PhD
Medical Director
Washington Adult Congenital Heart Program
Professor of Pediatrics
Children’s National Hospital
George Washington University

Thomas W. Carton, PhD, MS
Chief Data and Strategy Officer
Louisiana Public Health Institute

Keywords

Adult Congenital Heart Disease; Registry; Patient Engagement

Key Points

  • The research team reviewed the current challenges in adult congenital heart disease (ACHD) clinical care, surveillance, and long-term outcomes research, including a demand for subspecialty care that outweighs the supply; the heterogeneity of CHD and its long-term outcomes; and changing treatment strategies.
  • Though there are several CHD registries in the U.S., very few focus on adults with CHD. The research’s teams efforts, beginning with the Congenital Heart Initiative (CHI) launch in 2020, came about in response to this gap.
  • The mission of the CHI is to create a digital hub collecting health data from patients and providers. The team seeks to involve patients in every aspect of the CHI registry, with the ultimate goal of creating a community of ACHD patients and providers with a shared understanding of research needs and medical outcomes.
  • As the registry initially lacked data from healthcare providers and the electronic health record, the team partnered with PCORnet to establish an ACHD surveillance system.
  • The CHI-RON study uses PCORnet data and patient-reported outcomes to determine whether ACHD patients who are receiving recommended care do better than those who are not and the impact of gaps in care on patient-reported outcomes.
  • Dr. Carton provided an overview of PCORnet, followed by a review of the data science innovations the team used to generate a computable phenotype, identify congenital heart providers, diversify demographic recruitment for the registry, and incorporate procedure results into the common data model for analysis.
  • One of the strengths of this project was the high degree of patient and partner engagement. This involvement led directly to programming such as the “ACHA Café,” a virtual coffee hour in which patients could engage with one another, and guided their social media content.

Discussion Themes

The ACHD patient population identified a few key topics that they hoped the CHI would explore, including long-term effects of congenital heart defects, the most effective therapeutics for treating CHD, and mental health care options. Patient partner feedback also helped the team refine their recruitment and outreach materials.

Given the tendency for some types of CHD to be given more attention than others, the research team sought to enroll a patient population that was inclusive of a variety of presentations. This presented some methodological challenges, e.g. when it came to drawing distinctions in the registry.

Grand Rounds October 25, 2024: How Hybrid Working From Home Works Out (Nicholas Bloom, PhD)

Posted on by Sophia Ramirez

Speaker

Nicholas Bloom, PhD
Professor of Economics
Stanford University

Keywords

Remote Work; Hybrid Work; Job Performance; Attrition

Key Points

  • Dr. Bloom conducted a randomized controlled trial at the Trip.com headquarters in Shanghai to test the efficacy of a hybrid WFH model. 1,612 employees in the IT and Airfare divisions were randomized to come into the office either five days/week (the control arm) or three days/week, with the option to work remotely on Wednesdays and Fridays (the intervention arm).
  • The research team collected attrition and performance data during the six-month intervention period and 18-month follow-up period.
  • Over the six-month intervention period, quit rates were 33% lower in the treatment group than in the control group. The impact on attrition varied by group, with the greatest improvements among non-managers, women, and employees with long commutes. Satisfaction survey results were also more positive within the treatment group.
  • They found that hybrid WFH had no impact on performance reviews, promotion rates, or lines of code written. At the outset, managers were skeptical of the effect hybrid WFH would have on productivity; over the course of the trial, however, their attitudes became more favorable.
  • In the interest of reducing costs incurred by turnover, the company rolled out the hybrid WFH option to all employees shortly after the trial concluded. The research paper is open-access; read it in Nature.

Discussion Themes

To maximize the success of the hybrid WFH model, employees should coordinate the days that they come into the office with their colleagues and employers should base the number of days they allow people to work from home on their capacity to measure performance.

Though employees worked slightly fewer minutes on average on WFH days, taking advantage of the increased flexibility to pick up children, go to appointments, etc., they made up for that time by working more minutes on in-office days and on weekends.

There are tradeoffs involved in the decision to WFH: Though an employee may have more flexibility and even be more productive, they may also be less likely to get promoted compared to peers that are working in-person.

Going forward, Dr. Bloom predicts the percentage of hours worked from home will continue to increase due to three practical factors:

  1. Leases are expiring, and as they expire, companies will downsize to save money on real estate.
  2. Younger executives are more in favor of WFH.
  3. The technology is improving, making it easier and cheaper to WFH.

Grand Rounds October 18, 2024: Rigorous Testing of Behavior Change Interventions: Lessons from the BE ACTIVE Randomized Clinical Trial (Alexander Fanaroff, MD, MHS)

Posted on by Sophia Ramirez

Speaker

Alexander Fanaroff, MD, MHS
Assistant Professor of Medicine
Perelman School of Medicine
University of Pennsylvania

Keywords

Gamification; Financial Incentives; Physical Activity; Behavioral Economics

Key Points

  • Observational studies have found an inverse association between steps per day and risk for all-cause mortality and cardiovascular events. Though national surveys indicate the vast majority of people believe that exercise has many health benefits, many don’t exercise.
  • Behavioral economics seeks to explain why people make the decisions they do using concepts from economics and psychology. Dr. Fanaroff discussed how known biases in decision making – immediacy bias, status quo bias, and the endowment effect – create a disconnect between the “planner” self and the “doer” self. This framework can be used to understand why people don’t exercise, even when they know they should.
  • Another barrier that the research team identified was that existing guidelines from the American College of Cardiology and American Heart Association lack specificity and strong supporting evidence. Dr. Fanaroff ultimately introduced an alternative recommendation based on their findings.
  • The research team sought to leverage our existing biases in order to facilitate a sustained increase in physical activity. In a randomized controlled trial, they compared the effectiveness of gamification, financial incentives, and a combination of the two on increasing physical activity over a 12-month intervention period and a 6-month follow-up period.
  • After noting a discrepancy between the proportion of Black patients in the target population and participant population, the research team adjusted their recruitment strategy until they had a representative group. While direct-to-patient recruitment avoids biases related to who physicians invite to clinical trials, it may introduce biases related to who trusts researchers enough to join a study.
  • The research team found that gamification and financial incentives were roughly equally effective, leading to about 500 more steps per day than the control group. The combination arm saw the biggest improvement, with about 850 more steps per day than the control group. In practice, gamification is the more scalable and inexpensive of the interventions.
  • If stakeholders are to pay for an intervention like this, large scale clinical trials are needed to show improvements in patient-centered outcomes or clinical events.

Discussion Themes

Pharmaceutical interventions are not expected to be effective once treatment has ceased. If a behavioral intervention is shown to be effective during the period in which it’s administered, should it be held to a standard of long-term sustainability?

Patients set their own step count goals according to their baseline when the trial started.

The components of the intervention guided by behavioral economic theory, i.e. the immediacy of the innovation, the loss (rather than gain) of cash and/or points, and the support partners, differentiates it from the gamification built in to many wearables and fitness trackers.

Grand Rounds October 11, 2024: Early Diagnosis and Treatment of Asthma and COPD (Shawn Aaron, MD)

Posted on by Sophia Ramirez

Speaker

Shawn Aaron, MD
Professor
University of Ottawa

Keywords

Respiratory Symptoms; Guideline-Directed Care; Standard Care; Phone Recruitment; Randomized Controlled Trial

Key Points

  • Up to 70% of individuals with asthma or Chronic Obstructive Pulmonary Disease (COPD) remain undiagnosed. The 2007 – 2012 U.S. National Health and Nutritional Examination survey of randomly selected American adults found obstructive lung disease in 13% of the sample, 71% of whom had never been diagnosed.
  • Inspired by an encounter with a patient who had undiagnosed asthma, Dr. Aaron and his team developed three research questions for the Undiagnosed COPD and Asthma in the Population, or UCAP, study: 1) Can we find adults with undiagnosed asthma or COPD in the community? 2) Are they sick? 3) Can we treat them early to improve health outcomes?
  • They used a case-finding approach to identify symptomatic individuals with undiagnosed cases of either disease. Case-finding evaluates subgroups of people at increased risk of a disease; in this case, they looked at adults with undiagnosed respiratory symptoms.
  • To identify that sample, the team called over 1.1 million Canadian residents. After initial screening of the 50,000 contacts who indicated that there was someone in the household with respiratory symptoms, 2,857 of 4,272 eligible participants underwent testing with spirometry. A fifth of the sample had undiagnosed asthma or COPD.
  • Compared to a healthy age- and sex-matched control group, the adults with undiagnosed COPD or asthma had lower quality of life, worse symptoms and health status, and significant work impairment.
  • To determine whether early diagnosis of previously undiagnosed symptomatic asthma or COPD, and subsequent treatment, improves health outcomes, the team randomly assigned the participants diagnosed with asthma or COPD to the intervention or to usual care.
  • All participants and participants’ primary care providers (PCPs) were given a copy of their interpreted spirometry report with their diagnosis. The intervention group received treatment from a pulmonologist and an asthma/COPD educator; the control group received usual care from their PCP.
  • Guideline-directed treatment of undiagnosed COPD or asthma by a pulmonologist and an educator was found to improve healthcare utilization, symptoms, quality of life and lung function more than usual care.
  • In practice, not all patients can or will be treated by a lung specialist. The trial results indicated that the health of people with asthma or COPD will still improve if they are diagnosed and receive the usual care.

Discussion Themes

Cold-calling, though ultimately effective, was an expensive and inefficient screening and recruitment method. In the future, the team will attempt to drive people with respiratory symptoms to the study website by advertising in the community and on social media.

The research team considered a few iterations of the control arm, including a design in which the control group was informed of their diagnosis later on in the trial. They decided to compare guideline-directed treatment to usual care after reviewing the ethical considerations and the potential for inappropriate randomization.

Case-finding in a sample of people living in asthma or COPD “hotspots,” or in other high-risk populations, could increase the efficiency of the method but narrow the applicability of the findings.

Volunteer bias likely impacted the diversity of the sample; a disproportionate number of the volunteers were white (97%) and older, with an average age of 63.

Grand Rounds October 4, 2024: Health Trends Across Communities – A Novel Health System-Public Health Data Partnership (Tyler Winkelman, MD, MSc; David Johnson, MPH)

Posted on by Sophia Ramirez

Speakers

Tyler Winkelman, MD, MSc
Division Director, General Internal Medicine
Hennepin Healthcare
Co-Director, Health, Homelessness, and Criminal Justice Lab
HHRI

David Johnson, MPH
Health Informatics and Epidemiology
Program Manager
Hennepin County

Keywords

Electronic Health Record; Data Sharing; Public Health; Health Systems; Partnerships

Key Points

  • Collaboration across public health and health care is essential to developing actionable data for both sectors. Electronic Health Record (EHR) data can be used to fill the gaps in public health data and foster collaboration.
  • During the COVID-19 pandemic, it became clear that data infrastructure in the U.S. was underdeveloped. This made addressing COVID-19 challenging, is currently making addressing the overdose crisis challenging, and puts the country at risk for any future epidemics.
  • The Minnesota EHR Consortium (MNEHRC), formed in March 2020, facilitated collaboration between health systems in order to address gaps in COVID-19 data sharing and communication. They were able to develop the technical infrastructure to aggregate and share EHR data for real-time public health needs. Over time, the prioritization of data sharing for developing broader community health indicators became possible.
  • MNEHRC’s mission is to improve health by informing policy and practice through data-driven collaboration among members of Minnesota’s health care community. Dashboards are publicly available at www.mnehrconsortium.org.
  • Dr. Winkelman described how they built out a common data model at each of the MNEHRC health systems using Observational Medical Outcomes Partnership (OMOP), a common language for EHR data. OMOP was chosen because it’s open-source; it has a robust international online community; and some sites in the state had experience with OMOP, which helped with capacity building.
  • MNEHRC and Hennepin County’s Center for Community Health partnered to build Health Trends Across Communities (HTAC-MN), a unique data collaboration of health systems and public health agencies. They seek to develop comprehensive community health data infrastructure in Minnesota, ultimately strengthening community capacity to build healthy communities and promoting health equity.
  • Next steps for HTAC include developing and implementing processes to identify and prioritize new conditions; evaluating HTAC; and developing a plan for long-term sustainability.

Discussion Themes

Developing a central data model facilitated the collaboration.

Onboarding Federally Qualified Health Centers (FQHCs) to the consortium takes longer because of their internal capacity restraints. The team has had to be creative with figuring out how to onboard them; they are adding FQHCs in Hennepin County through EPIC affiliate agreements with Hennepin Healthcare and other sites through Minnesota’s quality measurement agency.

This is a new tool with a lot of potential, especially for the field of public health; researchers could use it to measure the impact of large-scale public health interventions. The HTAC team hopes that they’ll be able to further define the value that the data source can offer over the next few years.